The computational worm: spatial orientation and its neuronal basis in C. elegans

https://doi.org/10.1016/j.conb.2011.06.009Get rights and content

Spatial orientation behaviors in animals are fundamental for survival but poorly understood at the neuronal level. The nematode Caenorhabditis elegans orients to a wide range of stimuli and has a numerically small and well-described nervous system making it advantageous for investigating the mechanisms of spatial orientation. Recent work by the C. elegans research community has identified essential computational elements of the neural circuits underlying two orientation strategies that operate in five different sensory modalities. Analysis of these circuits reveals novel motifs including simple circuits for computing temporal derivatives of sensory input and for integrating sensory input with behavioral state to generate adaptive behavior. These motifs constitute hypotheses concerning the identity and functionality of circuits controlling spatial orientation in higher organisms.

Highlights

C. elegans exhibits two forms of spatial orientation behavior, klinokinesis and klinotaxis, both of which require computation of the time derivative of chemosensory information. ► The time derivative is computed partly at the cellular and partly at the network level. ► The output of chemosensory neurons modulates the probability of stochastic course-correction events in klinokinesis ► It is hypothesized that course corrections in klinotaxis require phasic sensory gating at the level of motor neurons.

Introduction

Most animals are able to orient their locomotion with respect to a variety of cues in the environment (Figure 1). Typical cues include local gradients, landmarks, compass bearings related to the Earth's magnetic field or the position of celestial bodies such as the Sun, and internalized maps of the environment [1]. These abilities  collectively known as spatial orientation behaviors  are of scientific interest because they are fundamental to survival and goal directed. In addition, spatial orientation behaviors are unusually well suited to investigations of sensorimotor integration in the nervous system because the sensory cues that signal the goal, and the motor performance required to attain it, are directly observable and easily quantified.

Despite the interest and tractability of orientation behaviors, our understanding of their neuronal basis remains rudimentary. It is notable, therefore, that considerable progress has been made in understanding the neuronal mechanisms of spatial orientation in the nematode Caenorhabditis elegans. C. elegans exhibits a diverse repertoire of orientation behaviors in response to a variety of gradients by which it locates food, mating partners, and habitat. These behaviors include chemotaxis to salts and other soluble compounds [2, 3••], chemotaxis to odorants associated with food and mates [4••, 5], thermotaxis to preferred temperatures [5], aerotaxis to preferred levels of oxygen [6], and aerotaxis to low levels of carbon dioxide [7]. In most of these behaviors, the gradient is sensed predominantly by one or more pairs of left–right symmetric neurons. These neurons contact the environment through a pair of sensory pores in the head. Although the right and left sensory pores are separated by 10 μm [8], the worm effectively samples the gradient at a single point in space. This is because the worm crawls on its right or left side which places the axis defined by the pores orthogonal to the plane of the substrate. The fact that the worm samples the environment at a single point entails that orientation behaviors are guided by either absolute stimulus strength or by the time derivative of sensory input as the animal moves through the environment. As discussed below, the physiology of C. elegans chemosensation supports the latter view.

C. elegans is an advantageous experimental system for the neuronal and theoretical analyses of behavior. The adult hermaphrodite has a compact nervous system of only 302 neurons, for which there exists an essentially complete anatomical reconstruction, the celebrated ‘wiring diagram’ of the worm [9••, 10•]. The genetic tractability of C. elegans together with recent technical advances in nematode neurophysiology  including patch clamp recording [11], calcium imaging [12], and optogenetics [13]  has accelerated the pace of research into the neuronal basis of behavior in this organism. C. elegans is also remarkably amenable to mathematical and computational modeling. The great majority of its neurons are morphologically simple, with just one or two processes which in most cases are unbranched [9••]. Neuronal processes are likely to be short relative to the predicted length constant of C. elegans neurons [11] suggesting that many C. elegans neurons are probably isopotential [11], or nearly so. Consequently, individual neurons can often be modeled as single electrical compartments. Consistent with their electrical compactness, and the absence of voltage gated sodium channels in the C. elegans genome [14], most C. elegans neurons probably do not fire classical, sodium-dependent action potentials [11, 15] and synaptic transmission between C. elegans neurons is likely to be graded [16]. Thus, the widely used formalism of real-time recurrent neural networks [17], which is formally equivalent to networks of single compartment neurons with graded synaptic input–output functions, is a better approximation of network-level function in C. elegans than in most other systems.

Two main behavioral strategies have been identified for spatial orientation behaviors in C. elegans. These are klinokinesis and klinotaxis, also known, respectively, as the pirouette [18, 19••, 20] and weathervane mechanisms in C. elegans [21••]. Both strategies were first identified in gradients of tastants such as inorganic salts and other soluble compounds [2]. As discussed below, essential neuronal elements of both strategies have been delineated.

Section snippets

Klinokinesis behavior

The klinokinesis strategy involves a biased random walk up the gradient (Figure 2a). In the laboratory, C. elegans orientation behaviors are mainly studied in the aqueous environment of the surface of a moist agar filled plate. Locomotory thrust is generated by snake-like undulations. These undulations occur in the dorso-ventral plane because worms crawl on their sides. In a uniform concentration of a chemoattractant, locomotion in C. elegans consists of periods of relatively straightforward

Neuronal mechanisms of klinokinesis

A candidate neural circuit for klinokinesis in gradients of tastants has been identified (Figure 2c). Although earlier modeling studies provided existence proofs for circuits of C. elegans-like neurons sufficient to compute dC(t)/dt at the network level [30, 31], subsequent neurophysiological experiments demonstrated that this quantity is actually computed jointly at the cellular and network levels [32••, 33]; for a possible exception to this rule, see Ref. [34]. Three left–right pairs of

Klinotaxis behavior

In the klinotaxis strategy, by definition, the animal's course is continuously corrected toward the line of steepest ascent up the gradient (Figure 3a). Klinotaxis is therefore a deterministic rather than a stochastic strategy. C. elegans steers by adjusting the angle of its head with respect to the rest of body [48, 49]. Course corrections in klinotaxis take place during the side-to-side movements of the worm's head that occur as part of normal undulatory locomotion in nematodes. Klinotaxis

Neuronal mechanisms of klinotaxis

Neuronal ablations together with anatomical reconstructions imply the circuit shown in Figure 3c. The ASE neurons provide almost all of the chemosensory input to the klinotaxis circuit [21••]. The off-cell, ASE-right, is the dominant member of the ASE class in that ablating it produces a strong deficit in klinotaxis, whereas ablation of the on-cell, ASE-left, has little or no effect unless combined with ablation of its sister cell [21••]. However, when both ASE neurons are ablated, the

Symmetry breaking in klinotaxis

A model of symmetry breaking in C. elegans klinotaxis based on the principle of phasic sensory gating [52, 53] has been proposed [50]. The model is based on two main assumptions. First, it assumes that undulations of the neck are produced by antiphasic synaptic inputs to dorsal and ventral neck muscle motor neurons (Figure 3c). Such inputs might originate from a central pattern generator, for example. Second, the model assumes a nonlinear relationship between motor neuron synaptic input and

Evolution of on–off coding

Why is on–off coding pervasive in C. elegans sensory neurons required for spatial orientation behavior? In the early visual system of vertebrates and insects, where on–off coding is well known [55, 56], it has been suggested that this coding scheme reflects a requirement for energy efficiency [57]. The alternative strategy, in which single neurons encode increases and decreases in the stimulus, requires high basal rates of activity to achieve a similar dynamic range. On–off coding can also be

Phylogenetic considerations

Spatial orientation has been investigated in a wide range of aquatic organisms in addition to C. elegans. These include prokaryotes, unicellular organisms, insect larvae, crustaceans, and fish. Despite this diversity in species, only two main types of sampling strategies have been identified: single-point sampling as described above for C. elegans, and stereo sampling in which the organism simultaneously makes use of a widely separated pair chemosensory structures. Comparison of orientation

Outlook

Spatial orientation behaviors in C. elegans are well suited to investigations of sensory processing and sensorimotor integration. The neuronal analysis of klinokinesis has yielded a simple circuit motif for computing the time derivative of sensory information that can be applied to other sensory modalities. Theoretical analysis of klinotaxis has provided a model of how representations of behavior are integrated with biological outcomes to generate adaptive responses. These findings lead to

References and recommended reading

Papers of particular interest, published within the period of review, have been highlighted as:

  • • of special interest

  • •• of outstanding interest

Acknowledgements

The author was supported during this research by grant R01-MH051383 from the National Institute of Mental Health. C. Derby provided useful discussion. S. Faumont, E. Heckscher, T. Lindsay, and K. McCormick provided critical comments on the manuscript.

References (75)

  • M.L. Sherman et al.

    Chemical orientation of brown bullheads, Ameiurus nebulosus, under different flow conditions

    J Chem Ecol

    (2001)
  • H. Schone

    The Spatial Control of Behavior in Animals and Man

    (1984)
  • S. Ward

    Chemotaxis by the nematode Caenorhabditis elegans: identification of attractants and analysis of the response by use of mutants

    Proc Natl Acad Sci U S A

    (1973)
  • C.I. Bargmann

    Odorant-selective genes and neurons mediate olfaction in C. elegans

    Cell

    (1993)
  • J.M. Gray et al.

    Oxygen sensation and social feeding mediated by a C. elegans guanylate cyclase homologue

    Nature

    (2004)
  • S. Ward et al.

    Electron microscopical reconstruction of the anterior sensory anatomy of the nematode Caenorhabditis elegans

    J Comp Neurol

    (1975)
  • L.R. Varshney et al.

    Structural properties of the Caenorhabditis elegans neuronal network

    PLoS Comput Biol

    (2011)
  • M.B. Goodman et al.

    Active currents regulate sensitivity and dynamic range in C. elegans neurons

    Neuron

    (1998)
  • R. Kerr

    Imaging Excitable Cell Activity in C. elegans

    (2002)
  • G. Nagel et al.

    Light activation of channelrhodopsin-2 in excitable cells of Caenorhabditis elegans triggers rapid behavioral responses

    Curr Biol

    (2005)
  • C.I. Bargmann

    Neurobiology of the Caenorhabditis elegans genome

    Science

    (1998)
  • S.R. Lockery et al.

    The quest for action potentials in C. elegans neurons hits a plateau

    Nat Neurosci

    (2009)
  • R.D. Beer

    Parameter space structure of continuous-time recurrent neural networks

    Neural Comput

    (2006)
  • J.T. Pierce-Shimomura et al.

    The fundamental role of pirouettes in Caenorhabditis elegans chemotaxis

    J Neurosci

    (1999)
  • D.B. Dusenbery

    Analysis of chemotaxis in the nematode Caenorhabditis elegans by counter current separation

    J Exp Zool

    (1974)
  • Y. Iino et al.

    Parallel use of two behavioral mechanisms for chemotaxis in Caenorhabditis elegans

    J Neurosci

    (2009)
  • T.A. Rutherford et al.

    Wave forms of C. elegans in a chemical attractant and repellent and in thermal gradients

    J Nematol

    (1979)
  • T. Hills et al.

    Dopamine and glutamate control area-restricted search behavior in Caenorhabditis elegans

    J Neurosci

    (2004)
  • C.I. Bargmann

    Chemosensation in C. elegans

    WormBook

    (2006)
  • W.S. Ryu et al.

    Thermotaxis in Caenorhabditis elegans analyzed by measuring responses to defined thermal stimuli

    J Neurosci

    (2002)
  • D.A. Clark et al.

    Temporal activity patterns in thermosensory neurons of freely moving Caenorhabditis elegans encode spatial thermal gradients

    J Neurosci

    (2007)
  • D. Ramot et al.

    Thermotaxis is a robust mechanism for thermoregulation in Caenorhabditis elegans nematodes

    J Neurosci

    (2008)
  • H.A. Zariwala et al.

    Step response analysis of thermotaxis in Caenorhabditis elegans

    J Neurosci

    (2003)
  • B.H. Cheung et al.

    Experience-dependent modulation of C. elegans behavior by ambient oxygen

    Curr Biol

    (2005)
  • N.A. Dunn et al.

    A neural network model of chemotaxis predicts functions of synaptic connections in the nematode Caenorhabditis elegans

    J Comput Neurosci

    (2004)
  • N.A. Dunn et al.

    Circuit motifs for spatial orientation behaviors identified by neural network optimization

    J Neurophysiol

    (2007)
  • H. Suzuki et al.

    Functional asymmetry in Caenorhabditis elegans taste neurons and its computational role in chemotaxis

    Nature

    (2008)
  • Cited by (62)

    • Neuronal, mathematical, and molecular bases of perceptual decision-making in C. elegans

      2019, Neuroscience Research
      Citation Excerpt :

      C. elegans shows navigational behavior in response to various odorants (Bargmann et al., 1993). They sense chemicals only at their nose-tip, a single point on their body, and therefore are thought to respond to temporal changes in chemical concentration, rather than directly to spatial gradients, for their navigation (Bargmann, 2006; Lockery, 2011). The main navigational strategy of C. elegans is known as the “biased random walk”.

    • Role of simulation models in understanding the generation of behavior in C. elegans

      2019, Current Opinion in Systems Biology
      Citation Excerpt :

      C. elegans uses a wide range of signals from the environment to find food, avoid harmful conditions, develop appropriately, and mate [111]. Spatial navigation in the worm involves the coordination of two behavioral strategies [72,112]: klinokinesis, a random biased walk based on regulating the frequency of turning that occurs in the stochastic switch between forward and backward movement; and klinotaxis, a gradual directed steering based on the modulation of dorsoventral head swings. The first of these strategies was studied in a series of experiments that characterized the rate of turning as a function of the change in concentration experienced by the worm [73].

    View all citing articles on Scopus
    View full text